35results about How to "Build quickly and efficiently" patented technology

The invention provides a real-time voice recognition system and method. The real-time voice recognition method includes the steps of: collecting current voice and current scene information of a current user; constructing a current scene language model corresponding to the current user; judging the type of the current scene according to the current scene information, and searching a static language model corresponding to the type of the current scene from static language models corresponding to different scene types which are constructed according to historical scene information; and calling a universal language model, and recognizing voice of the current user based on a hybrid and acoustic model of the universal language model, the searched static language model and the scene language model corresponding to the current user. The real-time voice recognition method constructs language models in an offline and online combination manner based on various scene information, and thus can effectively improve the accuracy rate of voice recognition.

A network-connected server for meeting initiation has software executing on the server from a non-transitory physical medium, the software providing a function allowing a user to schedule a meeting and to configure availability conditions, a function for informing scheduled participants of the scheduled meeting, a function for receiving indications of availability for the scheduled meeting from communication appliances associated with scheduled participants, and a function for connecting the communication appliances to communicate when the configured availability conditions are met.

The invention provides a method and equipment for neoantigen epitope prediction. The method for neoantigen epitope prediction includes (1), building a candidate epitope library; (2), building a prediction model based on a historical database, and (3) taking at least one subset of the candidate epitope library as an input variable of the prediction model to predict neoantigen epitope. The candidate epitope library is composed of differential expression peptides which are peptides with difference between samples having predetermined state and samples having no predetermined state. The historical database is composed of known epitopes related with the predetermined state. By the method for neoantigen epitope prediction, antigen mutation information can be determined according to gene level and protein level really, neoantigen epitopes capable of being bound with MHC are acquired effectively rapidly, and binding capacity of neoantigen epitopes and NHC can be predicted precisely. Besides, the method is easy and convenient and capable of saving manpower and material resources greatly.

The invention provides a visual map calibration method and device and a storage mediu. The method comprises that image data collected by a visual sensor is obtained, a movement track of the visual sensor is determined according to the image data, an initial visual map is generated based on the movement track of the visual sensor and the image data, and the initial visual map is calibrated based on a reference measurement map to obtain a first visual map. An indoor scene map serves as the reference measurement map. The visual map can be established rapidly and effectively in the calibration process, and accurate map data is provided for visual positioning.

A cross-modal hash learning method based on anchor map, characterized in that the method comprises the following steps: (1) using an objective function designed based on anchor map technology Binaryhash encoding of objects in image modal and text modal and the projection matrix of the image modal and text modal; (2) In view of the non-convex nature of the objective function, solving the unknownvariable in the objective function by alternate updating; (3) projection matrix based on the image modal and text modal obtained by the solution and generating a binary hash code for the sample of the query and the samples in the retrieved sample set; (4) Calculating the Hamming distance of the query sample to each sample in the sample set based on the generated binary hash code; (5) performing the retrieval of the query sample using the cross-modal retriever based on the approximate nearest neighbor search. The method can quickly obtain an approximate matrix of the true similarity matrix based on the anchor map technique.

The invention provides a method and device for constructing a three-dimensional terrain. The method includes the steps that satellite positioning data of a target zone on a map are extracted from a positioning log; the target zone is divided into grids in a preset plane so as to obtain grid nodes, in the preset plane, of the target zone; the elevation corresponding to the grid nodes is acquired according to the elevation of the extracted satellite positioning data; the three-dimensional terrain of the target zone is constructed in the preset plane based on the obtained grid nodes and the determined elevation.

When creating exhaustive, affected lists within a request scope, any legal risk due to failure to identify or leverage existing knowledge of a company about custody of evidence related to a legal matter is minimized. Accurate affected lists are constructed for an ongoing or impending litigation. Affected people, systems, or record types are included at first, and the list is modified over time once their exact association to the litigation context is clarified (new ones are added, not required are removed). Automation of data capture eliminates the risk of human error during data entry. Conflict resolution is applied when the same person, system, or record type is identified in the request scope, as part of multiple affiliations and associations. Overlapping lists are automatically merged to enable simple and efficient processing, without having to cull different lists. Affected elements in the request scope that could otherwise be missed are tracked and added. A relevant list of all affected elements based on multiple likely inconsistent sources of knowledge dispersed throughout a company is created, with a high degree of confidence in the context of an ongoing litigation or an impending litigation.

A building plenum is provided. The building plenum may include an assembly or housing and multiple building utility and safety infrastructures (such as fire, lighting, ventilation, and other building infrastructure). The building plenum may be constructed at a manufacturing facility for subsequent installation in a building. In this way, the building infrastructure may be quickly and efficiently installed in the building, such as by mounting the building plenum on a ceiling as part of a ceiling cover, ceiling tile, or the like. The building plenum may have different configurations (including different building utility and safety infrastructures), different lengths and widths, and different cover appearances (or other ornamentation).

The embodiment of the invention provides a voice extraction method, device and equipment. The method comprises the following steps: acquiring mixed voice sample data, wherein the mixed voice sample data comprises at least one of a noise signal, an interference voice signal and a reverberation signal and a target voice signal; training a preset voice separation model by using the mixed voice sample data to obtain a pre-trained voice separation model; constructing a strategy network and an evaluation network based on the pre-trained voice separation model, wherein the strategy network and the evaluation network correspond to network parameters; determining a target quantification strategy based on the network parameters; updating the pre-trained voice separation model by using a target quantification strategy to obtain a voice extraction model; and extracting a target object voice signal from to-be-processed voice data by using the voice extraction model. According to the method, the scale of the voice extraction model is reduced, so that the voice of the target object in the single-channel voice is quickly and effectively separated.

The invention discloses a plant mat ecological retaining wall used for restoring mine plants and a method thereof. The plant mat ecological retaining wall comprises a locating frame and plant mats, wherein the locating frame comprises locating anchor ropes, carrying frameworks, connecting hinges and locating fasteners, the carrying frameworks are connected with the plant mats through the locating fasteners, the plant mats are embedded in the carrying frameworks, one plant mat is arranged in each carrying framework, the carrying frameworks are coaxially distributed with the plant mats, and the plant mats comprise locating rods, flow guiding pipes, non-woven fabric outer layers, polyethylene polypropylene wrapping layers, planting base materials and plant seeds. The embodiment of the method comprises the following four steps: selecting a region, mounting equipment, carrying out cultivating operation and curing. On one aspect, a stable vegetation protection system can be rapidly and effectively built for various mine slope surfaces. On the other aspect, the efficiency of combining the vegetation protection system and the mine slopes can be effectively increased, and the biological connecting relationship between the vegetation system and the mine slopes is improved. Therefore, the efficiency of fixing and modifying mine operation surfaces is increased, and at the same time, the labor intensity and cost of vegetation curing operation are reduced.

The invention relates to a visual component model data construction method and system and electronic equipment. The scheme comprises the following steps: initializing model data; initializing filtering parameters; traversing the node data, and distributing the node ID of each node; establishing a set membership of nodes through connecting line data in the model data; traversing the node data; screening out correct root nodes and output nodes; analyzing and verifying all the nodes and adding analysis results to the corresponding nodes; storing and converting the successfully analyzed data and then returning the data to the front end; enabling the front end to verify the returned data; if verification fails, prompting the user to continue to modify the model; if verification succeeds, running the analyzed model. The method has the advantages that a multi-component, multi-data-source and large-data-volume query scene is supported, and model data constructed by the front-end component is correctly, efficiently and easily processed and analyzed into executable SQL statements.

Disclosed are a mesh topology generation method and a device based on an extensive makeup language and data format. The method comprises defines a data structure based on an extensive makeup language (XML). The topological information of all nodes is stored in the XML data structure. Connecting elements can be formed according to the topological information of all the nodes in the XML data structure. Each connection element comprises an initial node, a terminal node and a connected relation between the initial node and the terminal node. Each connection element can be verified and the topological structure can be generated according to the verifying results. The technical scheme of the mesh topology generation method and the device based on the extensive makeup language and data format can be fast to achieve a network topology structure and convenient to expand.

The invention provides a method for constructing a log object of a key operation, which comprises the following steps of obtaining the class types of objects in a data table so as to obtain all methodsets of the objects based on the class types; traversing the method set to obtain a method name of the method, and obtaining a write-in method for writing in a log table; intercepting the method nameto obtain a specific field name of the related attribute, and obtaining an obtaining method for reading the data value of the attribute from the data table through the field name; and executing an obtaining method to obtain the data value of the attribute, and writing the data value of the attribute into the log table through a writing method. Through the method, the log object can be quickly andeffectively constructed, and the maintenance cost is reduced.